Mortality Predicted by different Hematological Malignancy types, Bacteremia, and Lack of Chemotherapy at Kinshasa University Clinics

Objective To identify environmental, pollutant metals, epidemiological, clinical, therapeutic, and able to predict case fatality among patients with anemia and (HM). Methods This was a retrospective cohort study in black patient’s ≥ 20 years managed for recurrent fever, infections, anemia, and bacteremia between 2009 and 2015 at Kinshasa University Clinics, DRC. The outcomes such as incident HM and case fatality were assessed using univariate means, standard errors, relative risk (RR) and multivariate proportional Hazard ratio (HR) by Cox regression analysis, while Log-Rank test was performed for comparisons by Kaplan Meier curves (means, 95% CI, and median survival). Results Out of 105 incident HM patients (Male: Female ratio = 1), 57.1% (n= 60) experienced case fatality, There was no association between gender, residence, ethnicity, fever and mortality. There was a significant univariate association between age ≥ 55 years, serum chromium, serum cadmium, serum iodine, serum mercury, rainy season, hemolytic anemia, HM, bone pain, elevated Erythrocyte sedimentation, splenomegaly, abdominal pain, neutropenia, El Nino years, combined local dry season + global climate variability/El Nino, thrombocytopenia, multi-transfusions, bacteremia, lack of chemotherapy, and mortality. After adjusted for confounders using Cox regression models, only incident HM (HR = 16.8; 95% CI, P< 0.0001), Lack of chemotherapy (HR = 6.9, 95% CI, P< 0.002), and Bacteremia (HR= 4.2; 95% CI, P= 0.020) were the most significant and independent predictors of mortality. A separate analysis for HM patients, the mortality rates did not vary (P> 0.05) across HM sub-types: 71.4% (n= 10/14) in multiple myeloma, 84.3% (n= 15/18) in acute myeloid leukemia, 70% (n= 7/10) in myelodysplastic syndromes, 80% (n= 8/10) in chronic myeloid leukemia, 33.3% (n= 2/6) in acute myeloid leukemia, and 80% (n= 4/5) in acute lymphoid leukemia. from of socio-demographics, environmental factors, pollutant heavy metal toxicity, micronutrient deficiency, and mortality among with impact in term of clinical research, capacity building and prevention of such predictors, incidence, and fatality of HM.


Introduction
Cancer is well established as a mass problem in developed countries as well as in developing Sub-Saharan Africa countries (1). Hematological malignancies such as multiple myeloma (MM) leukemia, Hodgkin lymphomas, non-Hodgkin lymphoma (NHL) contribute around 10% of all cancers in Sub-Saharan Africa (2). Moreover, World Health Organization (WHO) report that developing countries are facing the threat of 50% person's death among HM patients (3).
In Sub-Saharan Africa, HM account for 9.9% of cancer deaths (2). Several environment factors (poverty), infections (HIV, Epstein-Barr virus, malaria, and Kaposi sarcoma associated herpes virus, inadequate health systems, lack of early diagnosis, chemotherapy, radiotherapy, stem cell transplantation, supportive care and palliation are associated with burden of mortality and mortality in HM patients from Sub-Saharan Africa (1).
Democratic Republic of Congo (DRC) Central Africa region, is facing emerging incidence of HM due to ageing, exposure to pollution (residence or occupation close to traffic, stations/garages, dust, smoke, or industrial areas) and combined local dry season + global climate variability/ La Nina and combined local dry season + global climate variability/El Nino (4). The most types of those Congolese HM are acute myeloid leukemia (AML), multiple myeloma (MM), myelodysplastic syndrome (MDS), chronic myeloid leukemia (CML), chronic lymphoid leukemia (CLL), and acute lymphoid leukemia (ALL) (4).
There is also significant association between heavy metal toxicity and HM patients from chaotic urbanization of Kinshasa megacity (5). However, there is no evidence on HM case fatality explained by biological, environment, clinical and pollutant heavy metals in Kinshasa, DRC. Sample size A random sample was calculated as follows: Ni = 4(Z α ) 2 P (1-P)/W 2 with Ni = sample size (Z α = Zvalue for two sided α of 0.05 and therefore a confidence interval of 95%, P = proportion of people with chronic anaemia expected to present with incident HM. Thus, the proportion of people with chronic anaemia who present with incident HM in the study population was not known, the authors assumed a proportion of 0.5 which will give the maximum sample size estimate (6), and a width or acceptable error (W) of 0.2. For a two sided α of 0.05 and therefore a confidence interval of 95%, zα = 1.96. So the sample size N = 4(1.96)2(0.5 × 0.5) / 0.04. N = 96 which was approximated to 100.
Allowance for missing data was made: 100 + 20% of potential misses = 120. Inclusion criteria were age ≥ 20 years and myelogram diagnosed HM. Patients with any myelogram result other than HM and those patients who did not wish to participate in the study were excluded.
Laboratory data HM was defined morphologically using 3.5 mL of blood obtained by venipuncture in tubes with anticoagulant ethylene diamin tetracetic acid (EDTA) for hemogram and myelogram.
Additional three smears of peripheral blood were undertaken. A sample taken from a tube with 3.8% citrate was used for the determination of the erythrocyte sedimentation rate (ESR). The morphological study used ten slides according to the May Grünwald Giemsa (MGG) stains and for the special stains (Sudan black B, Periodic Acid Shift, Coloring of Perls). Medullary smears were also performed for colored staining using the MGG viewed under the multi-ordinary microscope typical of Olympus. iFISH was performed on bone marrow (BM) and peripheral blood (PB) samples patients with chronic myeloid leukaemia (CML), acute myeloid leukaemia (AML) or acute lymphoid leukaemia (ALL) to confirm diagnosis for HM. Independent predictors were sex, age, heavy metals (Mn, Fe, Ni, Cu, Zn, As, Se, Br, Mo, Cd, I, Hg, Pb), HM types (MM, AML, MDS, CML, CLL, ALL).
Information on gender, age, exposure to traffic pollution and garages or stations, global climate variability (El Nino and La Nina), and local climate (dry and rainy seasons) was obtained. Climate changes caused by global warming conditions caused climate variability, which was defined as shortterm fluctuations around the mean climate state. Climate variability also refers to changes in climate patterns such as precipitation, weather conditions, temperature and humidity (7). El Nino Southern Oscillation is associated with climate changes in the tropical and sub-tropical regions as a result of temperature anomalies from the warming and cooling of the ocean surface (8,9). El Nino can be defined as warmer-than-normal sea or ocean surface temperatures (7) and La Nina refers to coolerthan-normal sea or ocean surface temperatures (7).

Statistical analysis
The reliability and validity of this research relies on the accuracy of data and the consistency of the tools and procedures used-the research design. In order to achieve the highest standard of accuracy, the researchers avoided confounding factors and foreseeable information bias and selection bias. In a univariate analysis, continuous variables were symmetrical and expressed as means ± standard deviation (SD), compared between two groups (HM and patients with anaemia but having a normal myelogram) using the Student's t-test. However, categorical variables were presented as frequencies (n = number) and prevalence (%), comparing the two groups using the chi-square test.
The researchers performed logistic regression and the Cox regression model to calculate adjusted multivariate hazard ratios (HR = beta exponential) for the risk of incident HM with their corresponding 95% confidence interval (95% CI). Kaplan-Meir curves after the Cox regression model produced a one minus cumulative survival function with a Log-Rank test of the equality of survival distribution and time medians for the different levels of stratified covariates. As age and laboratory optimal cut-off points were unknown, the effective (accurate and sensitive) cut-off values for discriminating HM and normal myelogram were tested a posteriori, using the Receiver Operating Characteristic curves (ROC) method. Diagnostic performance (prognostic) was characterized by the area under the curve (AUC for c-static) was calculated with its corresponding Standard Error (SE), 95% CI, sensibility and specificity.
The criterion for two-sided statistical significance was P-value <0.05. All analyses were performed using SPSS software version 23.0 for Windows (IBM/SPSS Inc., New York, USA).

Discussion
This is the first study in Sub-Saharan Africa to demonstrate burden mortality among Central African patients with HM confirmed by FISH. Demographic factors (aging, ethnic bantu from DRC provinces exposed to wars and ethnic conflicts (Central and East), bacteremia, severe anemia, lack of chemotherapy, increase in blood transfusion, accelerated ESR, loss of antioxidant capacitymicronutrient deficiency, exaggerated oxidative stress/pollutant metal toxicity, and combined local dry season + El Nino phenomenon. This study highlighted also the importance of FISH to confirm and to invalidate morphologically diagnosed leukemias. As expected, confirmed HM, dry season and El Nino phenomenon were the most significant predictors of survival in this prospective cohort.

HM and covariables as predictors of mortality
The present study identified the environmental, epidemiologic factors (demographic), private clinics, therapeutic and onco-hematologic able to predict mortality due to the HM. With this intention, a retrospective troop was led near the black patients Bantu old of more than 20 years and assumption of responsibility with the CUK for fever, infections, weakens and bacteremia. The incidence of HM and mortality due to the HM were evaluated by using the relative risk (RR) and the risk proportional multivariate (HR) by the regression of Cox, while the test Log-Rank was carried out for the comparisons by the curves of Kaplan-Meier.
The extent of specific mortality (lethality) was also described between various sub-types of HM.

Relative mortality dependent on the HM and other independent predictors
The incidence rate of mortality estimated at 57,1% in the present study between 2009 and 2015, was by far higher and quintuple than the incidence rates (12%) in 2012 (10). Certain studies underline the contribution of the deaths related to the HM estimated at 9,9% of the load of death among all the cases of cancers (HM and other neoplasms) diagnosed in the world (1,10,11,12). However, the probability of mortality among the patients reached of LAM estimated at 8/10 in this thesis was similar to that reported in the literature (19).
Climate change leads to over-exposure to ultraviolet rays and lower levels of vitamin D, resulting in decreased levels of immunity. In addition, chronic exposure to benzene products, lead and mercury in the environment increases the risk of HM. Climate change, environmental pollution, aging and anaemic hypoxia are associated with an imbalance in the anti-oxidant/oxidant state (oxidative stress), which increases the risk of HM incidence (20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30). In addition, Metal ions have been found to interact with cell components such as DNA and nuclear proteins, causing DNA damage and conformational changes that may lead to cell cycle modulation, carcinogenesis or apoptosis (31)(32)(33).
Selenium (Se) and zinc (Zn) are essential nutrients that are required for various biochemical and physiological functions such as oxidation-reduction reactions (34).
In some situations, iodine can have an antioxidant power 3 times that of vitamin C, one of the most important exogenous antioxidants. Therefore, any iodine deficiency will lead directly to the decrease or loss of its antioxidant capacity, especially in tissues rich in iodine (placenta, gastric mucosa, breast tissue) (35)(36)(37) Predictors of mortality in the HM The present study in agreement with the literature, recognized the incidence of HM, the bacteremia (18)

Clinical implications and perspective of public health
The present findings from interactions of socio-demographics, environmental factors, pollutant heavy metal toxicity, micronutrient deficiency, oxidative stress, and mortality among patients with HM, will impact DRC health system in term of clinical governance, research, capacity building and prevention of such predictors, incidence, and fatality of HM.

Limitations and strength
This is the first study to identify environmental, pollutant metals, epidemiological, clinical, therapeutic, and able to predict case fatality among patients with anemia and HM.
However, the study was limited to some degree to such extent of unknown predictors not evaluated.

Conclusion
HM is a major cause of morbidity and mortality with epidemic rates explained by lack of chemotherapy, and recurrent bacteremia among Bantu patients facing aging, climate change, rainy seasons, and lack of palliative care.

Conflict of interest and financial disclosures
There is no competing financial interests in relation to the work described http://www.wma.net/en/10 policies/b3/index.html).
between lethal and non-lethal HM cases.

Figure 2
Diagnose performance of optimal cutoffs of biomarkers for micronutrient deficiencies/loss of antioxidant capacity (serum zinc, iodine, selenium) and biomarkers for exacerbated oxidative stress/pollutant metal toxicity (serum cadmium, mercury, lead, uric acid, GGT) to discriminate fatal cases and non fatal cases of Hematological Malignancy.  Survival function according to global climate variability. References